Back to results
US 8,797,448 B2Granted
Rapid auto-focus using classifier chains, MEMS and multiple object focusing
AssigneeDigitalOptics Corporation Europe Limited
InventorsAdrian Capata, Stefan Petrescu, Petronel Bigioi, Peter Corcoran
Life of this patent
Estimating the term…
Abstract
A smart-focusing technique includes identifying an object of interest, such as a face, in a digital image. A focus-generic classifier chain is applied that is trained to match both focused and unfocused faces and/or data from a face tracking module is accepted. Multiple focus-specific classifier chains are applied, including a first chain trained to match substantially out of focus faces, and a second chain trained to match slightly out of focus faces. Focus position is rapidly adjusted using a MEMS component.
Claims20
A method comprising:
- acquiring data of a digital image which depicts one or more objects;
- determining a plurality of in-focus probabilities by applying a plurality of in-focus classifier sets to the data;
- wherein each in-focus probability indicates a likelihood that any of the one or more objects is depicted in focus in the digital image;
- determining an in-focus cumulative probability based on the plurality of in-focus probabilities;
- determining whether the in-focus cumulative probability is below an in-focus threshold;
- in response to determining that the in-focus cumulative probability is below the in-focus threshold, performing the steps of:
- determining a plurality of slightly-out-of-focus probabilities by applying a plurality of slightly-out-of-focus classifier sets to the data;
- wherein each slightly-out-of-focus probability indicates a likelihood that any of the one or more objects is depicted slightly-out-of focus in the digital image;
- determining slightly-out-of-focus cumulative probability based on the plurality of slightly-out-of-focus probabilities;
- and
- in response to determining that the slightly-out-of-focus cumulative probability is equal to, or above, a slightly-out-of-focus threshold, determining a particular object, of the one or more objects, that is depicted in the digital image slightly-out-of-focus;
- and
- wherein the method is performed by one or more computing devices.
10 more elements
A non-transitory computer-readable storage medium, storing one or more computer instructions which, when executed by one or more processors, cause the one or more processors to perform:
- acquiring data of a digital image which depicts one or more objects;
- determining a plurality of in-focus probabilities by applying a plurality of in-focus classifier sets to the data;
- wherein each in-focus probability indicates a likelihood that any of the one or more objects is depicted in focus in the digital image;
- determining an in-focus cumulative probability based on the plurality of in-focus probabilities;
- determining whether the in-focus cumulative probability is below an in-focus threshold;
- in response to determining that the in-focus cumulative probability is below the in-focus threshold, performing the steps of:
- determining a plurality of slightly-out-of-focus probabilities by applying a plurality of slightly-out-of-focus classifier sets to the data;
- wherein each slightly-out-of-focus probability indicates a likelihood that any of the one or more objects is depicted slightly-out-of focus in the digital image;
- determining slightly-out-of-focus cumulative probability based on the plurality of slightly-out-of-focus probabilities;
- and
- in response to determining that the slightly-out-of-focus cumulative probability is equal to, or above, a slightly-out-of-focus threshold, determining a particular object, of the one or more objects, that is depicted in the digital image slightly-out-of-focus.
8 more elements
A device, comprising:
- an image retrieval unit coupled to one or more memory units and configured to acquire data of a digital image which depicts one or more objects;
- and
- an image analysis unit configured to:
- determine a plurality of in-focus probabilities by applying a plurality of in-focus classifier sets to the data;
- wherein each in-focus probability indicates a likelihood that any of the one or more objects is depicted in focus in the digital image;
- determine an in-focus cumulative probability based on the plurality of in-focus probabilities;
- determine whether the in-focus cumulative probability is below an in-focus threshold;
- in response to determining that the in-focus cumulative probability is below the in-focus threshold, performing the steps of:
- determine a plurality of slightly-out-of-focus probabilities by applying a plurality of slightly-out-of-focus classifier sets to the data;
- wherein each slightly-out-of-focus probability indicates a likelihood that any of the one or more objects is depicted slightly-out-of focus in the digital image;
- determine slightly-out-of-focus cumulative probability based on the plurality of slightly-out-of-focus probabilities;
- and
- in response to determining that the slightly-out-of-focus cumulative probability is equal to, or above, a slightly-out-of-focus threshold, determine a particular object, of the one or more objects, that is depicted in the digital image slightly-out-of-focus.
10 more elements
Dependencies are inferred from the claim text.